Before an unproven process goes live, the backup systems, long-tenured staff, and informal channels a department relied on are removed as housekeeping, so the rollout's failure gets blamed on the department.
An example
In the quarter before a freight forwarder switched to an automated carrier-booking platform, its new head of operations retired the old booking system the booking desk had kept running in parallel, offered early retirement to two clerks who had handled exceptions for twenty years, and closed the shared inbox the desk used for quick answers from the claims team. Each step was logged as modernization. The platform went live in October. Exceptions the clerks used to catch from memory began reaching carriers late, and the desk had no fast route to claims. Within six weeks the company had paid late-booking charges on forty shipments and two large customers had complained. The head of operations told the executive committee the booking desk was not adapting to modern tools and proposed outsourcing it. Nobody had written down what the parallel system, the clerks, and the inbox had been doing.
Why it works
Stripping out the backup systems, long-tenured people, and informal channels a department relies on makes its normal work far more expensive just as a new, unproven process arrives, so failure becomes likely without any sabotage of the process. The department's record of competence looks worse in hindsight, as if the supports had done the work and the people had not. Removing them in advance also makes it cheaper to declare the department a failure later, since nobody has to explain away buffers that are already gone. It works when the supports are invisible to formal leadership, when the target lacks cover to say what was removed, and when the timeline leaves no room to rebuild.
The thinking behind it
Oliver E. Williamson Williamson's The Economic Institutions of Capitalism (1985) describes asset specificity: knowledge of one firm's systems or customers loses much of its value outside the relationship. Two clerks who caught exceptions from memory held highly specific knowledge that the new platform did not replace. Writing down what each informal support does, and what it prevents, makes that knowledge visible before anyone removes it.
Lee Ross The fundamental attribution error, which Ross named in The Intuitive Psychologist and His Shortcomings: Distortions in the Attribution Process (1977), is the observer's tendency to explain outcomes by people's traits and miss their situation. A booking desk whose backups and experienced staff were removed the quarter before go-live is judged unable to adapt. A public map of the removed buffers shows the situation the observers would otherwise miss.
Warning signs
In the quarter before an unproven rollout, legacy backup systems, long-tenured staff, and informal escalation channels are retired under a modernization or efficiency label. Each removal is small and defensible alone, and the department affected was never asked what it relied on. The rollout timeline leaves no time to rebuild what was removed. When problems start, they are put down to the department's failure to adapt.
How to defend yourself
Write the supports down first, since the move only works on ones nobody recorded.
- Keep a running list of the informal supports your department depends on: backup systems, people whose memory catches exceptions, favors and channels with other teams. Note what each one prevents.
- Get that list credited in planning documents and reviews, so removing any item requires a stated replacement.
- When a removal is proposed before a rollout, ask in writing what will do its job, and ask for the rollout date to move if nothing will.
- Find an auditor, an institutional historian, or a senior champion who can map the removed buffers in public.
Do this before the failure story sets. Afterward, the removed supports will already be on record as waste.
How a leader shuts it down
Before any rollout, require an inventory of what the affected department relies on, informal supports included, and a replacement for each one removed. Keep redundancy in place through the quarter an unproven change goes live. Manage organizational slack on purpose, since cutting it always looks like housekeeping.